Analysis of malignancy in pap smear images using gray level co-occurrence matrix and gradient magnitude

Hyperchromasia is one of the most common dysplastic change occur in cervical cell images particularly in the nucleus region. The texture of an image is a function of spatial variations of the gray level values and it is used to measure the variations of the pixel intensity of the surface in an image. Gray level co-occurrence matrix (GLCM) is widely used for texture analysis and it shows how repeatedly the different combinations of gray level values occur in an image. Textural changes in the region of interest can be identified by relating the co-occurring pairs of pixels spatially in various orientations and distance. Gradient magnitude and histogram equalization methods are used to enhance the details of the image. Biological variations in the textural region of the nucleus are analyzed and statistical significance at each point on the image is calculated. Various properties like energy, entropy, contrast, homogeneity, correlation and autocorrelation are computed for the classification of Pap smear cell images.